agentera 0.0.0 → 3.0.0-dev.1

package/bundle/skills/agentera/capabilities/orkestrera/schemas/validation.yaml

@@ -0,0 +1,107 @@
+VALIDATION:
+  1:
+    id: V1
+    rule: never_read_source
+    description: >-
+      Orkestrera MUST NEVER read implementation source code. It dispatches
+      capabilities; it does not implement. Artifact files (PLAN.md, HEALTH.md,
+      DECISIONS.md, PROGRESS.md, etc.) are not source code and are expected
+      to be read. This rule verifies the prose explicitly states the
+      source-code read prohibition.
+    severity: critical
+    checks:
+      - "instructions.md contains NEVER read implementation source code"
+  2:
+    id: V2
+    rule: evaluation_gate_required
+    description: >-
+      Every completed task MUST be verified by inspektera before being marked
+      complete. The orchestrator MUST NOT skip evaluation. This rule verifies
+      the prose contains the never-skip-evaluation safety rail.
+    severity: critical
+    checks:
+      - "instructions.md contains NEVER skip evaluation"
+  3:
+    id: V3
+    rule: retry_budget_enforced
+    description: >-
+      No task may be retried more than 2 times. After the second failure,
+      the task must be marked blocked. This rule verifies the retry budget
+      is stated in the prose and safety rails.
+    severity: critical
+    checks:
+      - "instructions.md contains retry budget of 2"
+  4:
+    id: V4
+    rule: exit_marker_required
+    description: >-
+      Every orkestrera invocation MUST emit an exit marker. The marker uses
+      the canonical glyph ⎈ (SG12) in the format ⎈ orkestrera · <status>
+      where status is one of EX1-EX4.
+    severity: critical
+    checks:
+      - "instructions.md contains exit marker with glyph ⎈ (SG12)"
+      - "instructions.md references exit signals EX1-EX4"
+  5:
+    id: V5
+    rule: orchestrator_read_only_artifacts
+    description: >-
+      The orchestrator reads artifacts and updates PLAN.md task statuses and
+      TODO.md blocked entries. It does NOT write to PROGRESS.md, CHANGELOG.md,
+      or any other capability-owned artifact. This rule verifies the
+      artifacts schema reflects this constraint.
+    severity: warning
+    checks:
+      - "PROGRESS.md is consumes only (not produces)"
+      - "HEALTH.md is consumes only (not produces)"
+      - "DECISIONS.md is consumes only (not produces)"
+      - "VISION.md is consumes only (not produces)"
+  6:
+    id: V6
+    rule: orchestration_context_first
+    description: >-
+      Orkestrera startup and evaluation MUST use
+      `agentera prime --context orkestrera --format json` and the
+      returned orchestration context before raw plan, progress, health, TODO,
+      or decisions artifact reads.
+    severity: critical
+    checks:
+      - "instructions.md contains agentera prime --context orkestrera --format json"
+      - "instructions.md contains Use the returned `orchestration_context` before raw plan, progress, health, TODO, or decisions artifacts"
+      - "instructions.md contains If `source_contract.complete_for_orchestration_context` is true, do not read raw plan, progress, health, TODO, or decisions artifacts"
+  7:
+    id: V7
+    rule: cli_fallback_before_raw_read
+    description: >-
+      When orchestration context is incomplete, Orkestrera MUST run listed
+      routine CLI fallback commands before any last-resort raw artifact read.
+    severity: critical
+    checks:
+      - "instructions.md contains run the listed routine CLI fallback commands"
+      - "instructions.md contains Read a raw artifact only as a last-resort diagnostic"
+  8:
+    id: V8
+    rule: caveats_preserved_for_evaluation
+    description: >-
+      Orkestrera MUST preserve compacted-decision, stale health/profile/app,
+      missing-state, and retry-state provenance caveats when handing a task to
+      inspektera.
+    severity: critical
+    checks:
+      - "instructions.md contains compacted decision caveats"
+      - "instructions.md contains stale health/profile/app caveats"
+      - "instructions.md contains retry-state provenance"
+  9:
+    id: V9
+    rule: satisfaction_authority_boundary
+    description: >-
+      When Orkestrera touches decision satisfaction, it may pass through or request
+      provisional satisfaction with evidence but MUST NOT user-confirm satisfaction.
+      Missing, compacted, open, provisional, or review-needed satisfaction state
+      remains a caveat and review pressure; automation must not reconstruct hidden
+      outcomes or claim it proved user intent.
+    severity: critical
+    checks:
+      - "Provisional satisfaction requires evidence"
+      - "Only the user confirms final satisfaction"
+      - "Missing or compacted satisfaction caveats are preserved"

package/bundle/skills/agentera/capabilities/planera/instructions.md

@@ -0,0 +1,368 @@
+# PLANERA
+
+**Planning Logic: Adaptive Notation, Executable Requirements Architecture. Enumerate, Refine, Assign**
+
+Scale-adaptive planning bridging deliberation and execution. PLAN.md with behavioral acceptance criteria for realisera. Planera owns WHAT and WHY; realisera owns HOW.
+
+Voice: the sharp colleague, here to plan the work. Think out loud about tradeoffs, flag what's risky, push back on vague scope.
+
+Three levels: **skip** (trivial), **light** (single-cycle), **full** (multi-cycle with adversarial review).
+
+Skill introduction: `─── ≡ planera · planning ───`
+
+---
+
+## Visual identity
+
+Glyph: **≡** (protocol ref: SG5). Used in the mandatory exit marker.
+
+---
+
+## State artifacts
+
+One file and one archive directory in `.agentera/`.
+
+| Artifact | Purpose | Bootstrap |
+|----------|---------|-----------|
+| `PLAN.md` | Canonical plan artifact, stored as `.agentera/plan.yaml` unless mapped otherwise. Spec, tasks, acceptance criteria. | Created during planning session. |
+| `.agentera/archive/` | Completed or discarded plans. | Created on first archival. |
+
+**Presence signal**: `.agentera/plan.yaml` means active planned work. Absence means no plan, so realisera reasons from VISION.md.
+
+For read-only current-plan startup or evaluation context, use `agentera plan
+--format json` first. When its `source_contract.complete_for_plan_artifact` is
+true, its `summary`, `entries`, and `source_contract` are complete for plan
+metadata, task dependencies, acceptance criteria, task evidence, overall
+acceptance, surprises, and previous-plan archive references; do not reread the
+persisted plan artifact defensively. Raw mapped plan artifact access is still
+required for writing a new plan, archiving a completed plan, artifact validation,
+corruption diagnostics, or when CLI output is unavailable or incomplete after
+CLI fallbacks.
+
+When a user directly invokes ≡ planera to create a new plan and the existing
+PLAN.md is already complete, archiving that completed plan before writing its
+replacement is implicit in the direct Planera invocation and does not require a
+separate pre-write confirmation. Human-initiated replacement plans still require
+plan approval before the write. Replacing, discarding, or archiving an active or
+incomplete plan is not implicit; ask for explicit confirmation before the write.
+
+Use `agentera describe --format json` and its `artifact_schemas` entry for `plan` to locate the active installed schema; use `artifact_locations` or `agentera query --list-artifacts --format json` to locate the mapped plan path and raw-access boundary before any diagnostic raw read. Do not search Agentera directories manually. Existing plan artifacts provide repository-local examples of the shape.
+
+For normal Planera startup, `agentera prime --context planera --format json`
+exposes `capability_context.context.planning_context.startup_contract` as the
+compact read model for planning levels, step markers, CLI-first orientation, plan
+artifact boundaries, full-plan review/self-audit requirements, and handoff
+expectations. Use that context before rereading this prose. Read this file when editing Planera,
+resolving contradiction or ambiguity, validating detailed behavior not covered
+by compact context, or investigating benchmark/read-trigger evidence.
+
+### Artifact path resolution
+
+Before reading or writing any artifact, prefer the CLI artifact-location contract from `agentera describe --format json` or `agentera query --list-artifacts --format json`. It resolves `.agentera/docs.yaml` mappings, default paths, existence, and read boundaries without a separate raw docs read. Raw `.agentera/docs.yaml` access is reserved for docs artifact writes, validation, corruption diagnostics, or unavailable/incomplete CLI discovery.
+
+### Contract values
+
+Contract values are inlined where referenced. Visual tokens from protocol: status tokens VT1-VT4 (■/▣/□/▨), list item VT15 (▸), inline separator VT16 (·), section divider VT14, flow/target VT17 (→). Skill glyphs SG1-SG12 for cross-capability references. Exit signals EX1-EX4 for the exit marker. Decision labels DL1-DL3 for DECISIONS.md reading. Severity issue levels SI1-SI4 for TODO entries. Phases PH1-PH5 for phase context.
+
+`references/contract.md` (at the v2 skill location `skills/agentera/references/contract.md`) remains available as a full-spec reference.
+
+---
+
+## Step 0: Detect level
+
+Assess work complexity. Read the description (user, DECISIONS.md, or TODO.md). Scan codebase if needed.
+
+| Signal | Level |
+|--------|-------|
+| Single-file change, bug fix, config tweak, < 50 lines | **Skip** |
+| One module affected, clear scope, fits one realisera cycle | **Light** |
+| Multiple modules, multi-file changes, 3+ logical steps, new feature spanning architecture | **Full** |
+
+**Skip**: This doesn't need a plan. Suggest ⧉ realisera and wait for
+confirmation unless the user already asked to implement now. Stop here.
+
+**Light or Full**: Proceed to planning.
+
+If uncertain between light and full, default to light.
+
+---
+
+Step markers: display `── step N/6: verb` before each step (Step 0 excluded). Steps: orient, specify, review, audit, write, handoff.
+
+## Step 1: Orient
+
+Use CLI-provided startup state and listed CLI fallbacks before raw artifact reads.
+For current PLAN.md context, prefer `agentera plan --format json`; if its source
+contract is complete, skip raw `.agentera/plan.yaml` reads during this read-only
+orientation step.
+
+Orient on VISION.md, DECISIONS.md, TODO.md, HEALTH.md, PROGRESS.md, and profile
+state through the supported CLI state seam where available. If a required state
+family is missing or incomplete, run the CLI-provided fallback before any
+last-resort raw artifact read.
+
+1. **VISION.md**: the north star (if exists)
+2. **DECISIONS.md**: use `agentera decisions --format json`; read `firm` (DL1) entries only (hard constraints for planning), and preserve returned `missing_fields`, `compacted`, `caveats`, and `satisfaction.review_needed` pressure instead of raw-reading missing historical context.
+3. **HEALTH.md**: latest codebase health grades (if exists)
+4. **TODO.md**: related known issues (if exists)
+5. **PROGRESS.md**: what was built recently (if exists)
+6. **Decision profile**: read `$PROFILERA_PROFILE_DIR/PROFILE.md` directly when it exists. If missing, proceed without persona grounding.
+
+7. **Project discovery** (if unfamiliar):
+   - Map directory structure
+   - Read dependency manifests, README.md, CLAUDE.md, AGENTS.md
+   - Identify build/test/lint commands
+
+Before decomposing: summarize the constraints from VISION.md and DECISIONS.md. If the decisions source contract reports `complete_for_normal_deliberation_context=true`, do not read raw `.agentera/decisions.yaml` for normal planning context.
+
+---
+
+## Step 2: Specify
+
+Define WHAT and WHY. Intent layer, not implementation details.
+
+Effort-bias check: when comparing plan shapes, do not treat effort spent constructing an option as evidence for it.
+
+### Light plans
+
+Brief conversation (2-3 questions):
+
+- **What**: one-paragraph description of the change
+- **Why**: what value it delivers or what problem it solves
+- **Constraints**: what must NOT break, what's out of scope
+- **Acceptance criteria**: 3-5 behavioral criteria in Given/When/Then format
+
+Write PLAN.md. Present for approval (human-initiated) or proceed (autonomous).
+
+### Full plans
+
+Deeper conversation:
+
+- **What**: detailed description
+- **Why**: motivation, user impact, relationship to VISION.md
+- **Constraints**: architectural boundaries, off-limits modules
+- **Scope**: what's in, out, deferred
+- **Design**: high-level approach. NOT implementation details.
+- **Task decomposition**: 3-8 ordered tasks, each one realisera cycle. Per task: description, dependencies, 3-5 behavioral Given/When/Then acceptance criteria
+- **Test proportionality**: for tasks with tests, add a proportionality target. Default: one pass + one fail per testable unit.
+- **Plan-level current-state check**: every full plan ends with a final state sync task depending on all prior tasks.
+- **Version bump check**: add a bump task when DOCS.md versioning exists and the plan includes `feat`/`fix` work.
+- **Overall acceptance criteria**: behavioral criteria for the complete feature
+
+Present for approval or proceed to adversarial review.
+
+---
+
+## Step 3: Review (full plans only)
+
+Spawn an adversarial critic. The critic MUST find issues.
+
+```
+You are reviewing a development plan for [project]. Your job is to find problems.
+
+## The plan
+[Full PLAN.md content]
+
+## Your mandate
+You MUST identify at least one issue. "Looks good" is not acceptable.
+
+Look for:
+- Tasks too large for a single implementation cycle
+- Missing dependencies between tasks
+- Acceptance criteria too vague to verify
+- Acceptance criteria that leak implementation details
+- Scope gaps or scope creep
+- Ordering issues
+- Conflicting constraints
+- Unacknowledged risks
+```
+
+Address legitimate issues; dismiss false positives with rationale. Present reviewed plan.
+
+---
+
+## Step 4: Pre-write self-audit
+
+Pre-write self-audit: run `agentera lint --artifact <ARTIFACT> --text "<DRAFT>"` (or `--file <PATH>`; schema names such as `decisions` auto-resolve the artifact file when no input is given) on the draft entry to check verbosity overruns, abstraction creep, and filler accumulation.
+Max 3 revision attempts. Flag with [post-audit-flagged] if still failing.
+
+## Step 5: Write PLAN.md
+
+Write tasks with acceptance criteria. The conversation preserves reasoning; the artifact preserves the plan.
+
+Write the plan to `.agentera/plan.yaml`.
+
+Artifact writing follows contract Section 24 conventions.
+
+### Light plan format
+
+```yaml
+header:
+  level: light
+  created: YYYY-MM-DD
+  status: active
+  title: Short Title
+what: One paragraph.
+why: Motivation and value.
+constraints: What must not break; what is out of scope.
+overall_acceptance:
+  - GIVEN context WHEN action THEN expected outcome
+tasks: []
+```
+
+### Full plan format
+
+```yaml
+header:
+  level: full
+  created: YYYY-MM-DD
+  status: active
+  reviewed: YYYY-MM-DD
+  critic_issues: "N found, N addressed, N dismissed"
+  title: Short Title
+what: Detailed description.
+why: Motivation, user impact, relationship to vision.
+constraints: Architectural boundaries and off-limits modules.
+scope:
+  included: []
+  excluded: []
+  deferred: []
+design: High-level approach, not implementation details.
+tasks:
+  - number: 1
+    name: Title
+    depends_on: []
+    status: pending
+    acceptance:
+      - GIVEN context WHEN action THEN expected outcome
+overall_acceptance:
+  - GIVEN context WHEN action THEN expected outcome
+surprises: []
+```
+
+---
+
+## Step 6: Handoff
+
+- **Single task**: suggest ⧉ realisera to execute and wait for confirmation.
+- **Full plan**: suggest ⎈ orkestrera to execute the entire plan and wait for confirmation.
+
+---
+
+## How realisera reads PLAN.md
+
+When PLAN.md has pending tasks, realisera's Step 2 changes:
+
+1. Read PLAN.md
+2. Find tasks with `Status: □` (VT3) pending whose dependencies are all `Status: ■` (VT1) complete
+3. Pick the first eligible task
+4. Use the task's acceptance criteria as exit conditions
+5. After committing, update task status to `■ complete`
+6. If unexpected, add to `## Surprises`
+7. If a task doesn't make sense, mark it `skipped`
+
+When all tasks complete:
+
+1. Archive PLAN.md to `.agentera/archive/plan-{date}.yaml`
+2. Delete `.agentera/plan.yaml`
+3. Realisera resumes vision-driven work selection
+
+---
+
+## Safety rails
+
+<critical>
+
+- NEVER include implementation details in PLAN.md. Planera owns WHAT and WHY. Realisera owns HOW.
+- NEVER write acceptance criteria that reference implementation. Use behavioral, domain-language criteria only.
+- NEVER produce more than 8 tasks in a full plan. If work requires more, split into sequential plans.
+- NEVER modify PLAN.md during a realisera cycle except to update task status and add surprises.
+- NEVER skip adversarial review for full plans.
+- NEVER auto-approve plans when human-initiated. Present for approval.
+- NEVER plan trivial work. If skip level, say so and stop.
+
+</critical>
+
+---
+
+## Exit signals
+
+Report one of these statuses at workflow completion (protocol refs: EX1-EX4).
+
+Format: `─── ≡ planera · <status> ───` followed by a summary sentence.
+For flagged, stuck, and waiting: add `▸` (VT15) bullet details.
+
+- **complete** (EX1): PLAN.md written and approved, adversarial review ran for full plans, handoff suggested.
+- **flagged** (EX2): Plan produced with caveats: critic issues dismissed, scope larger than ideal, or acceptance criteria not fully behavioral.
+- **stuck** (EX3): Cannot plan: description too ambiguous, context files contradict, or user declined approval with no revision path.
+- **waiting** (EX4): Feature not specified enough for acceptance criteria, or key architectural constraints unknown.
+
+---
+
+## Cross-capability integration
+
+Planera is the bridge between deliberation and execution.
+
+### Fed by ❈ resonera
+
+When resonera's deliberation concludes with a decision to build, the next step is ≡ planera. DECISIONS.md provides the "why" context.
+
+### Feeds ⧉ realisera
+
+PLAN.md tasks become realisera's work queue. Task acceptance criteria become cycle exit conditions. Realisera updates status and logs surprises.
+
+### Feeds ⎘ optimera
+
+When a plan includes optimization-shaped tasks, those tasks can be delegated to optimera.
+
+### Informed by ⛶ inspektera
+
+HEALTH.md findings can trigger remediation plans. Inspektera reveals structural issues; planera produces a plan to address them.
+
+### Informed by ♾ profilera
+
+Decision profile calibrates planning depth and pattern preferences.
+
+### Informed by ⬚ inspirera
+
+When inspirera recommends patterns or libraries, planera incorporates them into the plan's design section.
+
+### Reads ⛥ visionera output
+
+VISION.md provides the north star that planera reads during Orient.
+
+### Fed by ▤ dokumentera (docs-first workflow)
+
+In the docs-first workflow, dokumentera writes intent docs first, then planera decomposes them into tasks.
+
+### Reads ▤ dokumentera versioning conventions
+
+Planera reads the `versioning` block from DOCS.md. When the plan includes `feat`/`fix` work, planera appends a version bump task.
+
+---
+
+## Getting started
+
+### Planning a new feature
+
+1. ❈ resonera: deliberate on what to build and why
+2. ≡ planera: plan how to build it
+3. ⧉ realisera or ⎈ orkestrera: execute
+
+### Planning a remediation
+
+1. ⛶ inspektera: audit codebase health
+2. ≡ planera: plan fixes
+3. ⧉ realisera: execute
+
+### Mid-feature replanning
+
+If realisera logs multiple surprises in PLAN.md:
+
+1. Read the surprises section
+2. ≡ planera: reassess tasks
+3. Resume ⧉ realisera
+
+### Skipping the plan
+
+For trivial work, planera detects skip level and routes to ⧉ realisera directly.

package/bundle/skills/agentera/capabilities/planera/schemas/artifacts.yaml

@@ -0,0 +1,62 @@
+ARTIFACTS:
+  1:
+    id: A1
+    artifact_id: plan
+    local_role: produces_and_consumes
+    description: >-
+      Active plan with tasks, dependencies, and behavioral acceptance criteria.
+      Planera writes this during the planning session and may read existing
+      plan context before replacing or refining it.
+  2:
+    id: A2
+    artifact_id: plan_archive
+    local_role: produces
+    description: >-
+      Created when a plan is fully
+      executed or explicitly discarded by the user. Preserved for historical
+      reference.
+  3:
+    id: A3
+    artifact_id: vision
+    local_role: consumes
+    description: >-
+      Planera reads this during Orient to understand project direction and
+      align the plan with the vision.
+  4:
+    id: A4
+    artifact_id: decisions
+    local_role: consumes
+    description: >-
+      Planera reads firm (DL1) entries only as hard constraints for planning.
+  5:
+    id: A5
+    artifact_id: todo
+    local_role: consumes
+    description: >-
+      Planera reads this to identify related known
+      issues that the plan should address.
+  6:
+    id: A6
+    artifact_id: health
+    local_role: consumes
+    description: >-
+      Planera reads this to factor structural issues into the plan.
+  7:
+    id: A7
+    artifact_id: progress
+    local_role: consumes
+    description: >-
+      Planera reads this to understand what was built recently.
+  8:
+    id: A8
+    artifact_id: profile
+    local_role: consumes
+    description: >-
+      Planera reads this to calibrate planning depth and pattern preferences.
+  9:
+    id: A9
+    artifact_id: docs
+    local_role: consumes
+    description: >-
+      Planera reads this first to resolve project-local artifact mappings and
+      reads the versioning block for version bump task generation.

package/bundle/skills/agentera/capabilities/planera/schemas/exit.yaml

@@ -0,0 +1,33 @@
+EXIT_CONDITIONS:
+  1:
+    id: E1
+    condition: complete
+    description: >-
+      PLAN.md was written and approved (or skip level confirmed), the
+      adversarial review ran for full plans, and the handoff to realisera
+      or orkestrera was suggested.
+    exit_signal: complete
+  2:
+    id: E2
+    condition: flagged
+    description: >-
+      The plan was produced but with caveats: adversarial critic issues
+      dismissed rather than resolved, work scope larger than ideal, or
+      acceptance criteria could not be made fully behavioral.
+    exit_signal: flagged
+  3:
+    id: E3
+    condition: stuck
+    description: >-
+      Cannot plan because the work description is too ambiguous to
+      decompose, required context files contain contradictions, or the
+      user declined to approve the plan with no clear revision path.
+    exit_signal: stuck
+  4:
+    id: E4
+    condition: waiting
+    description: >-
+      The feature or change to plan has not been specified with enough
+      detail to produce acceptance criteria, or key architectural
+      constraints are unknown and cannot be inferred from the codebase.
+    exit_signal: waiting

package/bundle/skills/agentera/capabilities/planera/schemas/triggers.yaml

@@ -0,0 +1,34 @@
+TRIGGERS:
+  1:
+    id: T1
+    description: >-
+      Direct invocation by name or slash command. Matches when the user
+      explicitly requests planera for structured planning.
+    priority: high
+    patterns:
+      - "planera"
+      - "/planera"
+  2:
+    id: T2
+    description: >-
+      Planning requests. Matches when the user wants to create a plan,
+      break work down, or specify requirements before building.
+    priority: medium
+    patterns:
+      - "plan this"
+      - "write a plan"
+      - "break this down"
+      - "decompose this"
+      - "spec this out"
+      - "plan before building"
+  3:
+    id: T3
+    description: >-
+      Architectural planning signals. Matches when the user is asking how
+      to approach building something or when the work is too large for a
+      single development cycle.
+    priority: medium
+    patterns:
+      - "how should we build this"
+      - "multi-step feature"
+      - "this is too big for one cycle"

package/bundle/skills/agentera/capabilities/planera/schemas/validation.yaml

@@ -0,0 +1,61 @@
+VALIDATION:
+  1:
+    id: V1
+    rule: behavioral_acceptance_criteria
+    description: >-
+      All acceptance criteria MUST be behavioral (Given/When/Then format)
+      and MUST NOT reference implementation details (class names, API
+      endpoints, database tables, function names, line numbers).
+    severity: critical
+    checks:
+      - "Acceptance criteria use Given/When/Then format"
+      - "No implementation details in acceptance criteria"
+  2:
+    id: V2
+    rule: max_8_tasks
+    description: >-
+      Full plans MUST NOT exceed 8 tasks. If work requires more, it must
+      be split into sequential plans.
+    severity: critical
+    checks:
+      - "Task count ≤ 8 in full plans"
+  3:
+    id: V3
+    rule: adversarial_review_required
+    description: >-
+      Full plans MUST undergo adversarial review. The critic MUST find at
+      least one issue. "Looks good" is not acceptable. Skip level and
+      light plans are exempt.
+    severity: critical
+    checks:
+      - "Adversarial review ran for full plans"
+  4:
+    id: V4
+    rule: plan_final_state_sync
+    description: >-
+      Every full plan MUST end with a "final state sync" task that depends
+      on all prior tasks. This ensures aggregate plan-level CHANGELOG.md,
+      PROGRESS.md, and TODO.md updates.
+    severity: critical
+    checks:
+      - "Full plan has final state sync task as last task"
+  5:
+    id: V5
+    rule: test_proportionality_target
+    description: >-
+      Tasks that include writing tests MUST have a proportionality target
+      in their acceptance criteria. Default: one pass + one fail per
+      testable unit. Override only with explicit rationale.
+    severity: warning
+    checks:
+      - "Test tasks state proportionality target"
+  6:
+    id: V6
+    rule: skip_level_for_trivial_work
+    description: >-
+      Planera MUST detect skip level for trivial work (single-file change,
+      bug fix, config tweak, < 50 lines) and route directly to /realisera
+      without producing a PLAN.md.
+    severity: critical
+    checks:
+      - "Trivial work routed to /realisera without plan"